Reproductive success and fry production of the paiche or pirarucu

Aquaculture Research, 2011, 42, 815^822
doi:10.1111/j.1365-2109.2011.02886.x
Reproductive success and fry production of the paiche
or pirarucu, Arapaima gigas (Schinz), in the region of
Iquitos, Perú
Jesu¤s Nu¤n! ez1,2, Fred Chu-Koo3, Magali Berland1, Lamberto Are¤valo3, Ola¡ Ribeyro3,
Fabrice Duponchelle1,2 & Jean Franc" ois Renno1,3
1
IRD UR 175 CAVIAR, Montpellier, France
UNFV-FOPCA, Mira£ores, Lima, Peru¤
2
3
Instituto de Investigaciones de la Amazon|¤ a Peruana ^ IIAP, Programa para el Uso y Conservacio¤n del Agua y sus Recursos
(AQUAREC), Iquitos, Peru¤
Correspondence: J Nu¤n! ez, 351 RUE J.F. BRETON, BP 5095, Montpellier cedex 05, France. E-mail: Jesus.Nunez@ird.fr
Abstract
Arapaima gigas (paiche) is the largest scaled ¢sh
species living in the Amazon basin. Its biology is
both fascinating and misunderstood. In a context of
over¢shing, hence reduced natural populations,
aquaculture of a ¢sh with such interesting characteristics (large size, high growth rate, no intramuscular spines) is an important issue. The development
of farming production would also reduce the ¢shing
pressure on natural populations and allow restocking programmes in certain areas. To determine
what factors may in£uence the reproductive success
in captivity, data from breeding reports for 2007^
2010 were collected among ¢sh farmers in the region
of Iquitos. In parallel, we carried out physicochemical measurements in di¡erent ponds where
these paiches breed, and conducted personal interviews about the general ¢sh management conditions. The results show that reproduction occurs
throughout the year but with a higher intensity
during the rainy season. It also highlights farms
that have performed much better than others, but
no single factor except feeding level has been clearly
associated with reproductive success. The environmental control of reproduction in paiche, therefore,
remains partly mysterious. To deepen this study, we
recommend the systematic sexing of breeders, extending reproductive behavioural studies, and examining the limnological factors involved in fry
mortality.
r 2011 Blackwell Publishing Ltd
Keywords: arapaima, pirarucu, paiche, reproduction, peru, ¢sh culture
Introduction
Arapaima gigas (Schinz) is the largest scaled ¢sh
species living in the Amazon basin (Chu-Koo &
AlcaŁ ntara, 2009). It is an emblematic species of the
Amazonian ¢sh fauna, also known as the Paiche in
Peru and Pirarucu in Brazil. This species is very attractive for aquaculture in the Amazon region, owing
to its many advantages. Arapaima gigas has the best
growth rate among Amazonian cultivated ¢sh species, of 10^15 kg per year (Bard & Imbiriba, 1986;
Pereira Filho & Roubach, 2005; Nun! ez, 2009; Rebaza,
Rebaza & Deza 2010), and its ¢let, which is very
popular, does not have intramuscular spines. It also
accepts low water dissolved oxygen levels, due to its
obligatory aerial breathing. Finally, it has a carnivorous diet but is still £exible because it can accommodate consumption of dead ¢sh or ¢sh feed pellets.
However, natural populations have been subjected to
excessive ¢shing in the past decades for human consumption and ornamental ¢sh trading (Chu-Koo &
Tello, 2010), and already for more than 15 years,
A. gigas has become a very uncommon species in the
Amazon, except in some natural reserves (Pacaya
Samiria in Peru or MamirauaŁ in Brazil). The Paiche
is now in the CITES II list, which means that its trade
815
Aquaculture Research, 2011, 42, 815^822
Reproductive aspects of ARAPAIMA in captivity J Nu¤n! ez et al.
and use are strictly controlled. Scarce available information concerns mainly the reproductive biology
in the wild (Guerra, 1980; Saavedra Rojas, Quintero
Pinto & Landines Parra 2005); however, it is in the
interest of ¢sh farmers to develop a controlled production of this species (Pereira Filho & Roubach,
2005; Nun! ez, 2009) and one of the collateral expected e¡ects of Paiche domestication is a decreased
¢shing pressure on wild populations. Nevertheless,
many attempts have been made in recent decades to
develop its production in captivity (Bard & Imbiriba,
1986; Imbiriba,1991; Alcantara & Guerra,1992; Guerra, AlcaŁ ntara, Padilla, Rebaza, Tello, Ismin! o, Rebaza,
Deza, Ascon, Iberico, Montreuil & Limachi 2002;
Pereira Filho & Roubach, 2005; Saavedra Rojas, Quintero Pinto, Lopez Hernandez & Pezzato 2005), but this
task is made di⁄cult by the dispersed knowledge on
A. gigas biology in captivity, especially in crucial aspects like reproduction and feeding.
Paiche domestication faces the di⁄culty of gender
determination, which is virtually impossible from
external characteristics except during the reproductive period; otherwise, sex has to be determined in
blood or plasma samples by biochemical methods
using vitellogenin or sex steroid ratios (Chu-koo,
Dugue¤, AlvaŁ n Aguilar, Casanova Daza, AlcaŁ ntara
Bocanegra, ChaŁ vez Veintemilla, Duponchelle, Renno,
Tello & Nun! ez 2009). Moreover, the ovary is composed only by a single naked lobe or gymnovarium
(Godinho, Santos, Formagio & Guimaraes-cruz
2005; Saavedra Rojas, Quintero Pinto & Landines
Parra 2005; Nu¤n! ez & Duponchelle, 2009) that makes
the practice of canulation to determine oocyte development, induction of oocyte maturation and arti¢cial
fertilization almost impossible. As the production of
¢ngerlings actually relies only on spontaneous reproductions in ponds, the knowledge of the optimal
requirements for reproductive success is a prerequisite to substantially increase the ¢ngerling o¡er in
this region. Nevertheless, a noticeable development
of A. gigas production is expected as most of the
actual ¢ngerling production is destined for aquaculture instead of ornamental ¢sh trade as before
(Chu-Koo & Tello, 2010). This study aims at analysing
the reproductive data collected from A. gigas farmers
in the region of Iquitos between 2007 and 2010.
The results reported here represent the ¢rst attempt
to provide answers on the magnitude of a sustained
production level of ¢ngerlings with the current
knowledge and management conditions. It also
evaluates the variations of some physicochemical
and environmental factors that could promote or
816
inhibit A. gigas reproduction in controlled environments.
Material and methods
Fish origin
Between 2000 and 2007, a programme to assist ¢sh
farmers in the region of Iquitos enabled the Instituto
de Investigaciones de la Amazon|¤ a Peruna (IIAP) to
provide six paiche breeders to selected local ¢sh
farmers. Arapaima breeders had been bred previously
in ponds for 3 years in the IIAP’s facilities of Quistococha.
Fish reproduction supervising, mandated by law,
started in 2007 and ¢sh farmers had to report to local
authorities all events (births, fry harvest, etc.) that occurred in their farms, thus facilitating data collection. Some farms already owned paiches before the
assistance programme; they did not originate from
IIAP’s stock but contributed to some reproductive
events. Arapaima breeder weights varied approximately from 60 to 100 kg.
Fish rearing and feeding
The paiches are fed live ¢sh (forage ¢sh), dead ¢sh
(trash ¢sh from local market) and sometimes downgraded chicken or chicks, ¢sh or chicken viscera, ¢sh
feed pellets and even fruits, although ¢sh-based food
is predominant in every farm. The amount of food per
capita (calculated as kg food paiche biomass ! 1) varied greatly among farms, and most farmers could not
precisely quantify how much food they had distributed to ¢sh daily, monthly or even yearly. For this reason, we had to split the farms in three feeding levels:1,
small quantity (the forage ¢sh are distributed about
once a month); 2, average quantity (the ¢sh are fed
about once a week) and 3, large quantity (the ¢sh are
fed daily). In the ¢rst level, the food is essentially forage ¢sh consumed progressively by paiches. In the
other two levels, the food is mainly composed of trash
¢sh supplemented, or not, with the other cited foodstu¡s.
Physicochemical parameters
The physicochemical characteristics of pond waters
were measured within a weekly period for all farms
during the rainy season (February^March), using
the Lamotte AQ-2 kit (Drop Count Titration, Chester-
r 2011 Blackwell Publishing Ltd, Aquaculture Research, 42, 815^822
Aquaculture Research, 2011, 42, 815^822
town, MD, USA). The following parameters were
considered:
carbon dioxide (CO2): absent if values were
o1.25 mg mL ! 1 corresponding to the detection
limit, present if values were 41.25 mg mL ! 1;
nitrites (NO2 ! ): absent if values were
o0.05 mg mL ! 1 corresponding to the detection
limit; present if values were 40.05 mg mL ! 1;
ammonia (NH41): absent if values were
o0.02 mg mL ! 1 corresponding to the detection
limit; present if values were 40.02 mg mL;
s
pH and temperature were measured using a YSI
digital meter (Yellow Springs Instrument Company, Yellow Springs, OH, USA) at 15 cm from the
surface;
hardness (mg L ! 1 of CaCO3) levels o60 mg mL ! 1
are considered to be low;
transparency was determined by the Secchi disc
method.
Hydromorphological parameters
The environmental parameters susceptible to play a
role in the reproductive success of the paiche have
been recorded in all farms. Six environmental parameters have been studied: pond area; pond maximum depth; presence of vegetation on pond water
surface; shape of the shoreline (slow or steep slope);
soil type (clay, sandy clay or sandy-muddy) and water
(£ow through or stagnant water).
Reproductive aspects of ARAPAIMA in captivity J Nu¤n! ez et al.
composed of a few ponds of di¡erent sizes and
shapes, including small water bodies created in the
lower parts of small water streams that drain the
abundant rainfall in the area. Most of the farms have
running water almost all year round. A few have limited water £ow or even depend exclusively on rainfalls. The maximum depth is generally comprised
between 1.5 and 4 m (Table 1) with variable £oating
vegetation cover, and with shorelines generally covered by grass or other small plants, which were exceptionally steep.
The physicochemical parameters measured varied
within a small range (Table1) and were characteristic
of good quality and soft water. In two cases, pH values
attained relatively high levels (6.5 an 8.0), whereas
the majority of other farms had pH values around
5.0 and hardness never exceeded 25 mg L ! 1. These
values correspond to acidic soft waters typical of
these environments. Water transparency (Table 1)
was generally low (mean of 33 cm). This low transparency is also typical of local waters and sometimes
after rainfall the sediment content contributes to dramatically decrease water transparency. Carbon dioxide, NO2 ! and NH41 levels were, in most cases,
below the detection limits of the measurement kit
and never reached critical values for ¢sh, indicating
that rearing conditions were compatible with the
known species preferenda.
Reproductive activity
Field data collection and analysis
We went to di¡erent ¢sh farms, and there we requested to speak to the people who tend the paiches.
Then, we explained the reason for our presence, and
requested permission to make measurements of their
pond water quality. During the interviews, we enquired about the total number of paiches and the
food types and quantities distributed. During the visit, we also evaluated the total pond area, soil type and
water surface vegetation.
Data were analysed with the free statistical software R, version 2.6.2 and Statgraphics Plus (StatPoint
Technologies,Washington, DC, USA).
Results
Pond characteristics
The visited farms are located along the road between
Iquitos and Nauta (Fig. 1). Farms are generally
The reproductive activity of A. gigas was determined
during the period 2007^2010 on the monthly number of spawning events reported by the farmers.
Spawning generally occurred in a shallow area of
the pond in a small depression or nest dug by the
two brooders. Reproductions occurred all year
round, although a marked seasonality was apparent
(Fig. 2). The maximum number of spawning events
was observed between December and April (73 out
of 100 spawnings over the 4-year period), corresponding mainly to the rainy season, short photoperiod and higher temperatures (Fig. 3). During the dry
season (May^September), reproductive activity (Fig.
2) was lower (27 out of 100 spawnings).
Seasonality of fry production
The seasonality of fry production (Fig. 2) was more
pronounced than that of spawning activity. The
monthly fry production peaked during October^
r 2011 Blackwell Publishing Ltd, Aquaculture Research, 42, 815^822
817
Aquaculture Research, 2011, 42, 815^822
Reproductive aspects of ARAPAIMA in captivity J Nu¤n! ez et al.
Figure 1 Aerial view of the ‘Carretera Iquitos-Nauta’ Loreto, Peru. Points indicate the location of the main farms.
Table 1 Summary of some physicochemical and physical characteristics of Arapaima gigas breeding ponds in the region of
Iquitos, Peru (Iquitos-Nauta road)
Minimum
Maximum
Mean
(" SD)
Carbon
dioxide
Nitrites
Ammonia
pH
Hardness
(mg mL ! 1)
Transparency
(cm)
Temperature
( 1C)
Area
(m2)
Depth
(m)
2
22
8.85
6.24
0.05
0.15
0.089
0.121
0.2
3
1.06
0.82
4.25
7.5
5.47
0.86
7.0
25.0
15.4
4.32
7
124
33
30
27.5
32.4
29.9
1.46
500
8 000
2 923
2 142
1.5
4
2.12
0.82
March, and a very limited fry production was
observed during the dry season from April to
September, corresponding to longer days and lower
temperatures (Fig. 3). Patterns of fry production and
number of fry per spawning were quite similar from
818
April to November, but di¡ered strikingly from December to March (Fig. 2). During this second half of
the maximum reproductive activity period, the
monthly number of ¢ngerlings per spawning decreased remarkably to reach its minimum in April.
r 2011 Blackwell Publishing Ltd, Aquaculture Research, 42, 815^822
Aquaculture Research, 2011, 42, 815^822
Reproductive aspects of ARAPAIMA in captivity J Nu¤n! ez et al.
45
40
35
30
25
20
15
10
5
0
Jun
Jul
Aug
Sep
Oct
Cumul. fry / month
Nov
Dec
(x10–3)
Jan
Feb
Mar
Apr
May
Monthly spawnings
Monthly fry nr / spawning (x10–3)
Figure 2 Four-year (2007^2010) cumulated monthly number of Arapaima gigas spawnings (thick grey line), total fry
(solid dark line) and fry number per spawning (dotted dark line) of 16 farms along the Iquitos-Nauta road.Values must be
multiplied by 103 for total fry and fry per spawning.
T (°C)
35
760
30
740
Rainfall (cm)
720
20
700
15
680
660
5
MAY
APR
MAR
FEB
JAN
DEC
NOV
OCT
SEP
AUG
640
JUL
0
JUN
Nr of spawnings
10
Photoperiod (min. of light)
25
Figure 3 Monthly variations of rainfall (left axis, thick dark line), temperature (left axis, dotted dark line), cumulated
(2007^2010) number of spawnings of Arapaima gigas females (left axis, light dark line) and photoperiod (right axis, dotted
black line) in the region of Iquitos. Sources: IIAP (temperature and rainfall) and Bureau des Longitudes: http://www.
bureau-des-longitudes.fr (photoperiod calculations for Iquitos).
In terms of ¢ngerling production per spawning, the
best period corresponded to October^December;
nevertheless, the higher ¢ngerling production rate
was observed in February^March (Fig. 2).
Farm e⁄ciency on reproduction and fry
production
Total fry production, which was highly variable
among farms, reached over 13000 in certain farms
r 2011 Blackwell Publishing Ltd, Aquaculture Research, 42, 815^822
819
12000
1200
8000
800
4000
400
0
1
2
3
4
5
6
7
8 9 10 11 12 13 14 15 16
Farms
0
Fry / female / year
Aquaculture Research, 2011, 42, 815^822
Fry / spawning
Total fry
Reproductive aspects of ARAPAIMA in captivity J Nu¤n! ez et al.
Figure 4 Cumulated monthly fry number (left axis, open bars), fry per spawning (right axis, shaded bars) and fry per
female per year (left axis, black bars) of Arapaima gigas over a 4-year period (2007^2010) for 16 farms along the road
Iquitos-Nauta.
during the 4-year period, while being o500 in
others (Fig. 4) and was positively correlated with the
number of females (P 5 0.044) and the feeding level
(P 5 0.01). No signi¢cant correlation was found with
other factors. The number of fry per spawning over
the 4-year period (Fig. 4) was highly variable too
(208^1215), but was not correlated with feeding level
(P 5 0.06).
The total number of fry produced per female during the 4-year period (results not shown) ranged
from 106 to 4254, corresponding to 26^1063 ¢ngerlings per female and per year (Fig. 4) and this parameter was correlated with feeding level (P 5 0.02),
but not with other physicochemical or hydromorphological factors.
Discussion
During this study, spawning generally occurred in a
shallow area of the pond in a small depression or nest
dug by the two brooders, which corresponded to previous description in the wild (Bard & Imbiriba, 1986;
Castello, 2008). One of the most important ¢ndings of
this study is the almost continuous reproductive activity of A. gigas reared in pond conditions, while this
species is reported as an annual spawner with the
capability of few spawning events during the rainy
season in Peru (Guerra, 1980) and Brazil (Imbiriba,
1991). Nevertheless, even if these results demonstrate
that a residual reproductive activity persists during
the dry season, the overall survival of ¢ngerlings on
harvest, 1 month in average after spawning, is very
weak during the dry season as attested by the low
820
fry number per spawning observed in this study. This
situation can be explained by putative adverse ecological factors prevailing during the dry season in the
ponds, which induced higher mortalities in larval
and ¢ngerling stages. This means that even if the environmental conditions are still favourable enough
for ¢sh reproduction, the conditions are not optimal
for larvae and ¢ngerling survival. During the rainy
season, maximum ¢ngerling survival is observed
during the ¢rst months of the £ood period (October^
December), but survival decreased progressively with
rainfall decline. As no speci¢c study has been completed on plankton or other available food sources
for larvae feeding, it is di⁄cult to conclude on the predominance of the feeding factor over other environmental variables on larvae survival.
The lack of information on the real fecundity of females also hampers the interpretation of the data, as
the average number of spawned eggs may also be lower during the dry season. In the Iquitos-Nauta area,
the maximum number of fry collected 1 month after
hatching from a single spawning was 3941, which is
considerably less than the maximum fecundity evaluated by various authors (reviewed by Imbiriba, 94).
The reported absolute fecundities ranged from
58 000 to 86 000 maturing oocytes in wild ¢sh or a
maximum of 11000 juveniles in pond conditions in
Brazil (Bard & Imbiriba, 1986). From previous observations (Godinho et al. 2005; Nu¤n! ez & Duponchelle,
2009), the oocyte development of A. gigas is group
synchronous with three vitellogenic oocyte cohorts
at di¡erent stages of development in the ovaries of
maturing females. This means that the e¡ective fecundity per spawning is probably closer to one-third
r 2011 Blackwell Publishing Ltd, Aquaculture Research, 42, 815^822
Aquaculture Research, 2011, 42, 815^822
of the reported maximum fecundity or could be
equivalent to the oocyte number of the larger oocyte
diameter cohort. Even if one takes into account this
hypothesis (i.e. around 20 000 spawned eggs), the total number of hatched larvae is very far below this
¢gure in view of our observations during this 4-year
study. The probability that such high rates of eggs
and young stages mortality exist in a species providing elaborated bi-parental care seems unlikely. This
species may experience high rates of atresia just before spawning or poor egg survival during the embryogenesis stage, occurring in the nest at the bottom of
the pond, which might partly account for why estimated fecundities on ovaries in ¢nal maturation
stage are much higher than the observed number of
guarded fry. Clearly, additional work is needed to
shed light on this issue and to better understand the
reproductive potential of the A. gigas in captivity.
The number of fry produced per female and per year
has been positively correlated with the feeding level;
the farms where the food distribution was very low exhibited also a lower number of spawning. Among all
other factors studied, none appeared to play a key role
on the number of spawning or the total number of fry
produced. In particular, the presence or absence of
£oating vegetation seemed to have no in£uence in the
reproductive activity of A. gigas in captivity. Concerning NH41, the absence of correlation with the number
of spawning indicated that NH41 had no direct e¡ect
within the measured range on female reproductive activity, and did not a¡ect ¢sh behaviour becauseA. gigas
can tolerate much higher concentrations than those
observed in this study (Cavero, Pereira-¢lho, Bordinhon, Fonseca, Ituassu¤, Roubach & Ono 2004).
Regardless of the number of spawning events, the
quantity of fry produced per brood varied greatly
from one farm to another. These di¡erences can be
explained by lower fecundities or di¡erential mortalities during the1-month average period before ¢ngerling harvest. The most likely hypothesis to explain
these variable mortalities is that feeding of fry depends greatly on the productivity of ponds. As juveniles and ¢ngerlings feeding in the wild consists
mainly of zooplankton, insects, insect larvae and
small gastropods (Oliveira, Poleto & Venere 2005),
the main causes of mortality are probably an insu⁄cient food availability. This hypothesis needs to be
tested by a detailed study of the limnological parameters and plankton productivity in the di¡erent
farming environments. If con¢rmed, considering the
huge variations observed on harvested fry among
farms, ¢ngerling production could be improved by
Reproductive aspects of ARAPAIMA in captivity J Nu¤n! ez et al.
properly fertilizing the ponds to allow optimal primary and secondary production.
The results described in this study emphasize the
high impact of feeding level on reproductive activity
and probably also on the survival of young stages and
it appeared that in most farms the feeding level of the
breeders was insu⁄cient. The importance of nutritional aspects has been also reported by (Saavedra
Rojas, Quintero Pinto, Lopez Hernandez et al. 2005)
for the quantity and quality of fry produced in captivity, but this is the ¢rst experimental observation of direct in£uence of the feeding level on reproductive
activity of A. gigas. In the Iquitos-Nauta area, feeding
of A. gigas is achieved mainly with frozen or live ¢sh,
this feed appeared to be e¡ective for breeding stocks
at 2% of biomass per day, although a more detailed
study is needed to establish the exact amount and
quality of food that would allow for optimal reproductive rate and fry survival in this species.
In conclusion, this study shows that under good
management conditions including su⁄cient food
supply, the production of A. gigas ¢ngerlings in captivity can be substantial (more than 1000 fry per female per year) and holds a good potential for
improvement, despite the low fecundity of this species. These results also substantiate the possibility of
sustained fry production almost during 8 months
year round from September to April. Finally, our
study con¢rms the possibility of aquaculture development of A. gigas through an improvement of
broodstock management.
Acknowledgements
This study was carried out as part of the scienti¢c
collaborations between the IIAP and the Institut de
Recherche pour le De¤veloppement (IRD), both part
of the research network known as RIIA (Red de Investigacio¤n sobre la Ictiofauna Amazo¤nica, http://
www.riiaamazonia.org).
The authors wish to thank all the farm owners and
people who contributed to the ¢eld work.
This work has been funded in part by the Peruvian
research programme INCAGRO.
This is a publication IRD-DIVA-ISEM 2011-042.
Reference
Alcantara F. & Guerra H. (1992) Cultivo de paiche, Arapaima
gigas, utilizando bujurqui, Cichlasoma bimaculatum, como
presa. Folia Amazo¤nica 4, 129^139.
r 2011 Blackwell Publishing Ltd, Aquaculture Research, 42, 815^822
821
Reproductive aspects of ARAPAIMA in captivity J Nu¤n! ez et al.
Bard J. & Imbiriba E.P. (1986) Piscicultura de pirarucu, Arapaima gigas. No. 52. EMBRAPA-CPATU, Bele¤m, Brazil pp. 52.
Castello L. (2008) Nesting habitat of Arapaima gigas (Schinz)
in Amazonian £oodplains. Journal of Fish Biology 72,
1520^1528.
Cavero B.A.S., Pereira-Filho M., Bordinhon A.M., Fonseca
F.A.L., Ituassu¤ D.R., Roubach R. & Ono E.A. (2004) Tolera“ncia de juvenis de pirarucu ao aumento da concentrac# a! o
de amoŒnia em ambiente con¢nado. Pesquisa Agropecua?ria Brasileira 39, 513–516.
Chu-Koo F. & AlcaŁ ntara F. (2009) Paiche dome¤stico en la
Amazon|¤ a: perspectivas de una crianza sostenible. Pesca
Responsable 57, 32^33.
Chu-Koo F. & Tello S. (2010) Produccio¤n de semilla de Paiche
en Peru¤. Infopesca Internacional 41, 30^35.
Chu-Koo F., Dugue¤ R., AlvaŁ n Aguilar M., Casanova Daza A.,
AlcaŁ ntara Bocanegra F., ChaŁ vez Veintemilla C., Duponchelle F., Renno J.F., Tello S. & Nun! ez J. (2009) Gender
determination in the Paiche or Pirarucu (Arapaima gigas)
using plasma vitellogenin, 17b-estradiol, and 11-ketotestosterone levels. Fish Physiology and Biochemistry 35,
125^136.
Godinho H.P., Santos J.E., Formagio P.S. & Guimaraes-Cruz
R.J. (2005) Gonadal morphology and reproductive traits
of the Amazonian ¢sh Arapaima gigas (Schinz,1822). Acta
Zoologica 86, 289^294.
Guerra F.H. (1980) Desarollo sexual del paiche, Arapaima
gigas, en las zonas reservadas del estado (R|¤ o Pacaya Samiria) 1971^1975. In : Informe IMARPE No. 67, pp. 20.
Instituto del Mar del Perú, Callao, Perú.
Guerra H., AlcaŁ ntara F., Padilla P., Rebaza M.,Tello S., Ismin! o R., Rebaza C., Deza S., Ascon G., Iberico J., Montreuil V.
& Limachi L. (2002) Produccio¤n y manejo de alevinos de
paiche. : IIAP, Iquitos, Peru¤.
822
Aquaculture Research, 2011, 42, 815^822
Imbiriba E.P. (1991) Produc" a! ao e manjo de alevinos de pirarucu,
Arapaima gigas (Cuvier). Technical Report EMBRAPA-CPATU, 57,19pp.
Nun! ez J. (2009) Domestication de nouvelles espe'ces d’inte¤reŒt
aquacole en Amazonie. Cahiers Agriculture 18,136^143.
Nu¤n! ez J. & Duponchelle F. (2009) Towards a universal scale
to assess sexual maturation and related life history traits
in oviparous teleost ¢shes. Fish Physiology and Biochemistry 35,167^180.
OliveiraV.D., Poleto S.L. & Venere P.C. (2005) Feeding of juvenile pirarucu (Arapaima gigas, Arapaimidae) in their natural environment, lago Quatro Bocas, Araguaiana-MT,
Brazil. Neotropical Ichthyology 3, 312^314.
Pereira Filho M. & Roubach R. (2005) Pirarucu, Arapaima
gigas. In Espe¤ cies nativas para piscicultura no Brasil (ed.
by H. Rosa Nascimento), pp. 37–62. UFSM, Santa
Maria, Brasil.
Rebaza M., Rebaza C. & Deza S. (2010) Densidad de siembra
para cultivos de Paiche en jaulas £otantes. Aquavisio¤n 6,
26^27.
Saavedra Rojas E.A., Quintero Pinto L.G. & Landines Parra M.A.
(2005) Aspectos reproductivos. In: Biolog|¤ a y cultivo del pirarucu¤ Arapaima gigas Schinz, 1822. Pisces: Arapaimidae.
Bases para un aprovechamiento sostenible ^ Aspectos Reproductivos (ed. by A.I. Sanabria, I.C. Beltran & P.V.
Daza), pp. 31–40. INCODER/UNC, Bogotá, Colombia.
Saavedra Rojas E.A., Quintero Pinto L.G., Lopez Hernandez
N. & Pezzato L.E. (2005) Nutricio¤n y alimentacio¤n del pirarucu¤ Arapaima gigas (Schinz, 1882). In Biolog|¤ a y cultivo
del pirarucu¤ Arapaima gigas Schinz,1822. Pisces: Arapaimidae. Bases para un aprovechamiento sostenible ^ Aspectos
Reproductivos (ed. by A.I. Sanabria, I.C. Beltran &
P.V. Daza), pp. 41–58. INCODER/UNC, Bogotá,
Colombia.
r 2011 Blackwell Publishing Ltd, Aquaculture Research, 42, 815^822